Making people safer from the dangers of roadkill might just be one of the least discussed elements in the advent of connected and autonomous vehicles.
Although no one officially collects data on the number of animal-vehicle collisions, it’s clear that it is a big problem. Virginia alone spends $4 million each year to remove the bodies of animals, according to Andy Alden of Virginia Tech. And that is just one measure of the cost. Vehicle-animal conflicts, as they are somewhat antiseptically referred to, kill drivers, passengers, and animals and cost car owners a pretty penny as well.
Andy Alden
Largely seen as a rural problem, increasing deer populations in urban areas indicate that solutions are needed that work in a variety of settings.
Mobility Momma interviewed Alden about ways sensing technology for automated and connected vehicles might help. As Alden points out, people are animals as well, and these innovations could make biking and walking safer too.
What are you considering at Virginia Tech that might reduce crashes with animals?
Alden: First, we are looking at detection technology on two different tracks – sensors in the infrastructure and sensors in the vehicle. In each instance, the first element is to detect animals and the second is to find what intervention will avoid a crash between the animal and the car.
Where does the technology for detecting animals via sensors in the roadway come from?
As part of a study for the Virginia Department of Transportation (VDOT), we have adapted a Department of Defense technology for perimeter detection for use on the roadside. In the first phase of this research, we tested it on our smart-road facility where we have deer, bear, bobcat, and coyote “intrusions.” Using a cable buried nine inches deep at the edge of the roadway, we could reliably detect these animals’ presence 95 percent of the time. Now we are putting in place a test on an active roadway, Route 8 between Floyd and Christiansburg.
Vehicle manufacturers have recently touted the use of sensors in cars and trucks to detect animals. How does that work?
As a prequel to autonomous driving, auto manufacturers and technology firms are looking at ways sensing technology – including cameras, lidar, and radar – can detect pedestrians, bicyclists, and now animals. In research done for Toyota, we used vehicles equipped with sensors as mobile-data detection units. Over a five-month period across seasons, we collected naturalistic data with the help of 48 participants driving in areas of incidents of animal-vehicle conflicts. Through that data, we were able learn about animal behavior – where do the animals cross, how fast they are traveling, their reaction to headlights, which direction do animals move in relation to the vehicle, and whether the animals are traveling in groups or individually.
Okay, so how do you stop crashes once animals are in the roadway?
To be determined. The sensor just detects the presence of an animal. We have looked at a variety of ways to prevent the crash. A roadside sensor could trigger a flashing light. I think drivers either won’t be inclined to respond to the flashing light or will not have enough time to react once they see the light. Vehicle manufacturers are looking at onboard sensors being able to trigger crash avoidance either through steering or braking.
At Virginia Tech, we have thought about triggering fencing that would stop the animal. Another approach would integrate detection in an app such as Waze or a safety app that would audibly notify drivers of all hazards – not just animals but debris, crashes, and work zones.
We have even thought about dispatching a drone to shoo an animal that is lingering on the roadway. VT also has a effort under way to use drones as an emergency-assistance tool.
Could the roadside technology be used in urban areas?
A roadside sensor is not a particularly appropriate approach in congested urban areas, where there would be conflicts with curbside activities, including sidewalks and parking. There may be suburban areas that have high incidences of animal crossings that would work.
Could this technology help protect humans?
I could imagine some special-use cases in which the roadside sensors could be deployed in high-risk zones as another tool to make a crossing safer for a blind pedestrian. Generally, vehicle-mounted sensors are more promising for giving visibility to bicyclists and pedestrians. Vehicle sensors could give truck and bus drivers 360 visibility of bicyclists and pedestrians and stop trucks and buses from crossing paths.
Photo by Jay Galvin/Flickr.